Component IV - Contingency Management Treatment Duration Contingency management (CM) interventions are highly efficacious in improving substance abuse treatment outcomes, but few studies have implemented this approach with alcohol use disorder patients. Further, no known studies have experimentally evaluated how duration of CM affects outcomes. This issue appears to be of central importance for impacting long-term behavior change, given the well-established association between length of treatment engagement and outcomes. In this application, we propose to evalua the efficacy of prize-based CM when administered according to a usual duration of 12 weeks versus an extended duration of 24 weeks. We will also investigate how probabilities of reinforcement may impact the relationships between CM duration and outcomes. Alcohol abusing or dependent patients (N=310) beginning intensive outpatient day treatment at three community based clinics will be randomly assigned to one of four conditions: (a) standard treatment as usual (ST) at the clinic without CM; (b) ST with CM for 12 weeks with a 0.5 probability of winning prizes for each negative sample submitted and an expected average maximum earnings of $300 in prizes; (c) ST with CM for 24 weeks with a 0.34 probability of winning prizes for each negative sample submitted and an expected average maximum earnings of $300 in prizes; or (d) ST with CM for 24 weeks with a 0.5 probability of winning prizes for each negative sample submitted and an expected average maximum earnings of $500 in prizes. During weeks 1-12, two breath samples per week will be collected from all patients, and those in the CM conditions will have the opportunity to win prizes for submission of negative samples. Those assigned to the longer duration CM conditions will continue to receive reinforcement for negative samples provided weekly during weeks 13-24. In group d, probabilities of reinforcement will be identical to those used in group b, but th will remain available for an additional 12 weeks. In group c, the probabilities of winning prizes will be lower so that the magnitude of overall reinforcement is consistent with group b. Alcohol use, other drug use, psychosocial problems, and HIV risk behaviors will be measured at baseline, during and post treatment, and throughout an 18-month follow-up period. We expect that CM will decrease alcohol use to a greater extent tha non-CM treatment, and that availability of CM for 24 weeks may result in longer term benefits than 12 week exposure to CM. This study will be the first to evaluate the effects of probability of winning prizes on response CM. We will also assess patient characteristics that may be associated with improved outcomes within and across treatments.